Background: Adults with intellectual disabilities should participate in the diagnostics of their mental health. The Glasgow Depression Scale for people with a Learning Disability (GDS-LD) and its Carer Supplement (GDS-CS) were translated into German and tested. Methods: Internal consistency, criterion validity and inter-test reliability were tested in 64 adults with borderline, mild or moderate intellectual disabilities and their carers. Convergent validity was analysed in 57 adults without intellectual disabilities. Results: Internal consistency was good ([alpha] = 0.81) for GDS-LD and acceptable ([alpha] = 0.72) for GDS-CS. The GDS-LD did not differentiate between groups with and without depression. The GDS-CS significantly differentiated between these groups. Convergent validity of the GDS-LD was high. The correlation of GDS-LD and GDS-CS was non-significant. Conclusions: In its present form the German version of the GDS-LD does not meet the psychometric properties to be used in clinical practice. This leads to the broader question, how to measure depression in people with learning disabilities with the knowledge of the fallibility of existing measures when utilised with this clinical cohort. Also, future studies need to investigate the role of self-rating.